3ftfbsdi sujdmf $ibsbdufsj[bujpopg$bsejbd …downloads.hindawi.com/journals/bmri/2013/503047.pdf ·...

Hindawi Publishing CorporationBioMed Research InternationalVolume 2013, Article ID 503047, 15 pageshttp://dx.doi.org/10.1155/2013/503047

Research ArticleCharacterization of Cardiac-Resident Progenitor Cells ExpressingHigh Aldehyde Dehydrogenase Activity

Marc-Estienne Roehrich,1 Albert Spicher,1 GiuseppinaMilano,2 and Giuseppe Vassalli1, 3

1 Department of Cardiology, Centre Hospitalier Universitaire Vaudois (CHUV), Avenue du Bugnon, 1011 Lausanne, Switzerland2Department of Cardiovascular Surgery, Centre Hospitalier Universitaire Vaudois (CHUV), Avenue du Bugnon,1011 Lausanne, Switzerland

3Molecular Cardiology Laboratory, Fondazione Cardiocentro Ticino, Via Tesserete 48, 6900 Lugano, Switzerland

Correspondence should be addressed to Giuseppe Vassalli; [email protected]

Received 16 August 2012; Revised 29 October 2012; Accepted 2 November 2012

Academic Editor: Franca Di Meglio

Copyright © 2013 Marc-Estienne Roehrich et al.is is an open access article distributed under theCreativeCommonsAttributionLicense, which permits unrestricted use, distribution, and reproduction in anymedium, provided the originalwork is properly cited.

High aldehyde dehydrogenase (ALDH) activity has been associated with stem and progenitor cells in various tissues. Human cordblood and bonemarrowALDH-bright (ALDHbr) cells have displayed angiogenic activity in preclinical studies and have been shownto be safe in clinical trials in patients with ischemic cardiovascular disease. e presence of ALDHbr cells in the heart has not beenevaluated so far. We have characterized ALDHbr cells isolated frommouse hearts. One percent of nonmyocytic cells from neonataland adult hearts wereALDHbr. ALDHvery-br cells weremore frequent in neonatal hearts than adult. ALDHbr cells weremore frequentin atria than ventricles. Expression of ALDH1A1 isozyme transcripts was highest in ALDHvery-br cells, intermediate in ALDHbr

cells, and lowest in ALDHdim cells. ALDH1A2 expression was highest in ALDHvery-br cells, intermediate in ALDHdim cells, andlowest in ALDHbr cells. ALDH1A3 and ALDH2 expression was detectable in ALDHvery-br and ALDHbr cells, unlike ALDHdim cells,albeit at lower levels compared with ALDH1A1 and ALDH1A2. Freshly isolated ALDHbr cells were enriched for cells expressingstem cell antigen-1, CD34, CD90, CD44, and CD106. ALDHbr cells, unlike ALDHdim cells, could be grown in culture for morethan 40 passages.ey expressed sarcomeric 𝛼𝛼-actinin and could be differentiated along multiple mesenchymal lineages. However,the proportion of ALDHbr cells declined with cell passage. In conclusion, the cardiac-derived ALDHbr population is enriched forprogenitor cells that exhibitmesenchymal progenitor-like characteristics and can be expanded in culture.e regenerative potentialof cardiac-derived ALDHbr cells remains to be evaluated.

1. Introduction

Growing evidence suggests the adult heart may harborresident stem and progenitor cells that participate in adaptiveresponses to myocardial injury, and possibly in cellularhomeostasis under normal conditions. In most studies, theprospective isolation of cardiac stem cells has relied uponthe use of speci�c antibodies that recognize cell-surfaceantigens expressed by stem cells in other tissues, particularlyby hematopoietic stem cells (HSCs). Cardiac stem cellsexpressing stem cell antigen-1 (Sca-1) or the stem cell factorreceptor, c-kit (CD117), have been described [1, 2]. A distinctpopulation of cardioblasts expressing the transcription factor

islet-1 (Isl-1) and entering fully differentiated cardiomy-ocyte lineages has also been identi�ed [3, 4]. However,the expression of cell-surface epitopes can vary with themetabolic state of the cell and the experimental conditionsused. erefore, no individual cell-surface antigen markerde�nitely identi�es a single entity of cardiac stem cell.Functional properties of tissue-resident stem cells that couldbe used for their prospective isolation, regardless of cell-surface marker expression, have been intensely searched for.It has been proposed that, irrespective of their lineal origin,stem cells may share common mechanisms to regulate self-renewal and differentiation, and that candidate “stemness”genes may serve as universal stem cell markers.

2 BioMed Research International

Cai et al. [5] proposed that high aldehyde dehydrogenase(ALDH) activity is one of a small set of common character-istics shared by stem cells among tissues (recently reviewedby Balber [6]). Few ALDH-bright (ALDHbr) pluripotentialcells, measured by Alde�uor stain, gave rise to all somaticand reproductive cell lineages in tunicates [7, 8]. ALDH isa cytosolic enzyme responsible for the oxidation of intra-cellular aldehydes. It plays important roles in oxidation ofalcohol and vitamin A, and in chemoresistance to cyclophos-phamide. ALDH superfamily is highly conserved acrossspecies [9]. Nineteen ALDH isoforms are known in human.ALDHbr, side scatter-low (SSClo) populations from murineandhumanbonemarrow (BM) [10–13], umbilical cord blood(UCB) [14–19], cytokine-mobilized peripheral blood [20],and circulating blood cells [21, 22] are enriched in stemand progenitor cells. e ALDHbrSSClo population includesvirtually all human CD34+ and CD133+ cells that establishlong-term, multilineage hematopoietic colonies in cultureor long-term, multilineage xenogras in immunode�cientmice [16]. More recently, ALDHbr stem cells have beenidenti�ed in nonhematopoietic systems, such as circulatingendothelial progenitor cells (EPCs) [21, 22], neural stem cells[23, 24], human muscle precursor cells with high myogenicactivity [25, 26], colonic stem cells [27], and mammary stemcells [28]. Moreover, ALDHbr cancer stem cells have beendescribed in multiple types of cancer and shown to predictpoor clinical outcomes in various contexts [27–29].

We therefore asked whether the postnatal heart harborsALDHbr cells, as measured by Alde�uor stain. We foundthat the neonatal as well as the adult mouse heart harbors arelatively small population of ALDHbr cells (≈1% of all non-myocytic cells present in the heart), which is enriched for cellsexpressing Sca-1 and other progenitor cell markers. ALDHbr

cells, but not ALDHdim cells, adhered to plastic and grew inculture. e expanded population expressed sarcomeric 𝛼𝛼-actinin, a cardiac marker, along with mesenchymal stem cell(MSC) markers. It could be induced to differentiate alongmultiplemesenchymal cell lineages.ese results suggest thatALDH may mark a population of cardiac-resident, MSC-like progenitor cells that possess superior ex vivo growthcharacteristics.

2. Methods

2.1. Mice and Cell Isolation. Neonatal (postnatal day 1),young adult (8 week-old), and aging (24 month-old) C57Bl/6mice were purchased from Charles River Laboratories(France). Immediately aer the sacri�ce of the mice, thechest was opened, a canula was introduced into the leventricular cavity, an incision was made in the right atrialwall, and the heart was perfused with heparinized PBS. eexplanted heart was washed in PBS and cut into small piecesthat were then placed in a Falcon tube (50mL) contain-ing 5mL RPMI 1640 medium (Invitrogen, Carlsbad, CA,USA) supplemented with 12.5𝜇𝜇L Liberase Blendzyme 4 and25 𝜇𝜇L DNAse I (both from Roche, Basle, Switzerland), andincubated for 45min at 37∘C under gentle shaking. e cell

suspension was �ltered through a 70𝜇𝜇m-�lter using a syringeplunger, washed in PBS, and centrifuged. e pellet wasresuspended in 100 𝜇𝜇L Dead Cell Removal Kit microbeadssolution (Miltenyi, Bergisch Gladbach, Germany), incubatedfor 15min at RT, changed to Dead Cell Removal Buffer, andpassaged through a Miltenyi LS column. e �ow-throughwas centrifuged and the pellet was resuspended in FACSbuffer. In a subset of experiments, atria and ventricles wereprocessed separately tomeasure ALDHbr cells in the differentcardiac chambers of origin.

2.2. Flow Cytometric Analyses. Cells were washed in PBS,resuspended in Alde�uor buffer, and then reacted withAlde�uor substrate (STEMCELL Technologies, �ancouver,BC, Canada) according to the manufacturer’s instructions.Aer the ALDH enzyme reaction, cells were washed, resus-pended in cold Alde�uor buffer, and maintained on iceduring all subsequent manipulations. A forward scatter(FSC) versus side scatter (SSC) cytogram was used to gatesignals from cells, and an Alde�uor �uorescence versus SSCcytogram was constructed. A unique ALDHbr populationwas present in Alde�uor-reacted samples and absent in cellsin the samples treated with the speci�c inhibitor of ALDH,diethylaminobenzaldehyde (DEAB). Cells incubated withAlde�uor substrate andDEABwere used to establish baseline�uorescence of these cells and to de�ne the ALDHbr andthe ALDHvery-br region as less than 0.1% and 0% of totalevents, respectively. Cell incubation with Alde�uor substratein the absence of inhibitor induced a shi in FL1 �uorescencede�ning the ALDHdim, the ALDHbr, and the ALDHvery-br

population. To assess cell-surface antigen expression, cellswere incubated with antibody for 20min, washed, and resus-pended in cold Alde�uor buffer. Flow cytometric analyseswere performed on a FACSCalibur instrument (BectonDick-inson, San Jose, CA, USA) operating at 488 nm excitationwith standard emission �lters. Alde�uor �uorescence wasmeasured in FL1 and APC in FL4. Gates used to resolveantigen-expressing cells were set using appropriate isotype-matched control Abs. Data �les containing at least 2,000ALDHbr cells were acquired for analysis using the CellQuestsoware (Becton Dickinson).

2.3. Immunostaining. Culture-expanded ALDHbr cellsgrown in Lab-Tek chambers were �xed with 1%paraformaldehyde (PFA) for 10min, followed by blockingsolution (1x PBS, 1% BSA, 0,3% Triton) for 1 h, andincubated with rat anti-mouse c-Kit/CD117 mAb (clone3C1; Miltenyi) coupled to APC for 1 h at RT, followed bygoat anti-rat IgG coupled to Alexa 488 (1 : 400 dilution;Molecular Probes, Life Technologies, Grand Island, NY,USA). To detect type II collagen, cells were �xed in PFA,incubated in blocking solution for 1 h, and then with mousemonoclonal anti-type II collagen 4 Abs coktail (1 : 100dilution; Chondrex, Redmond, WA, USA), followed bygoat anti-mouse IgG coupled to Alexa 488 (1 : 400). Todetect sarcomeric 𝛼𝛼-actinin, cells were �xed with 4% PFAfor 1 h, incubated in blocking solution for 1 h, and thenwith mouse antisarcomeric 𝛼𝛼-actinin mAb (1 : 750 dilution;

BioMed Research International 3

clone EA-53, Sigma-Aldrich, St. Louis, MO, USA) at 4∘Covernight, followed by goat anti-mouse IgG coupled to Alexa488 (1 : 400). To detect 𝛼𝛼-smooth muscle actin (𝛼𝛼-SMA),cells were stained with anti-𝛼𝛼-SMA rabbit polyclonal Ab(abcam5694; Abcam). Control sections were incubated withsecondary Ab only. Nuclei were stained with Hoechst 33342and slides were mounted with Mowiol. Pictures were takenwith a Zeiss Axioplant 2 microscope (100x objective).

2.4. Antibodies. e following mAbs were used to detectthe cell-surface antigen distribution of cardiac cells by �owcytometry: CD11b-APC (clone M1; Miltenyi), CD14-APC(clone Sa2-8; eBioscience; San Diego, CA, USA), CD29-PE (clone RTK2758; BioLegend), CD31-APC (clone 390;eBioscience), CD34-APC (clone RAM-34; eBioscience),CD38-APC (clone 90; eBioscience), CD40-APC (clone1C10; eBioscience), CD44-APC (clone IM7; eBioscience),CD45-APC (clone 30-F11; eBioscience), CD90.2-APC(clone HIS51; eBioscience), CD105-PE (clone MJ7/18;eBioscience), CD106-PE (clone 429; eBioscience), c-Kit/CD117-APC (clone 3C1; Miltenyi), CD133-APC(clone 13A4; eBioscience), CD140b-APC (eBioscience),CD146-FITC (Miltenyi), Flk-1-APC (clone Avas12a1;eBioscience), Lineage Cell Detection Cocktail-Biotin mouse(Miltenyi; cat. no. 130-092-613), MHC class II-APC (cloneM5; eBioscience), NG2 chondroitin sulfate proteoglycan(Chemicon/Millipore; cat. no. AB5320), Sca-1-APC (cloneD7; eBioscience), rat IgG2a isotype control-APC (cloneeBR2a; eBioscience), rat IgG2a isotype control-FITC (cloneeBR2a; eBioscience), rat IgG2a isotype control-PE (cloneeBR2a; eBioscience), rat mouse IgG1, 𝜅𝜅 isotype control(clone P3; eBioscience), and Arminian hamster IgG isotypecontrol (clone eBio299Arm; eBioscience).

2.5. Real Time RT-PCR. Freshly isolated ALDHdim, ALDHbr,and ALDHvery-br cells from PBS-perfused hearts from 8-week-old mice (𝑛𝑛 𝑛 𝑛) were sorted using a Beckman CoulterMoFlo Astrios FACS system. Total mRNA from each cellsubset was extracted using the RNeasy Micro kit (Qiagen).e different c-DNAs were generated using the Quantitectreverse transcription kit (Qiagen) from total mRNA obtainedwith a genomic DNA digestion step according to the man-ufacturer’s instructions. cDNA (1 : 10 dilution) was used forquanti�cation using the RT2 SYBRGreen qPCR Kit (Qiagen)and the Rotor-Gene 2000 system (Qiagen) according tothe manufacturer’s instructions. Real-time PCR reactions(in triplicates) were set up in 10 𝜇𝜇L reaction volume with5𝜇𝜇L of RT2 SYBR Green mix, 0.4 𝜇𝜇L of 10 𝜇𝜇M RT2 qPCRPrimer Assay, 3 𝜇𝜇L cDNA, and 1.6 𝜇𝜇L water. e polymerasewas heat-activated for 10min at 95∘C, and the reactionswere then cycled 50 times (95∘C, 15 sec; 55∘C, 40 sec; 72∘C,30 sec), followed by a melting step. Primers were obtainedfrom (Qiagen). Relative expression was calculated with thecomparative ΔCt method using GUSB as a reference gene.

2.6. Ex Vivo Cultures of 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑏𝑏𝑏𝑏 Cells. In two prelimi-nary experiments, cells from enzymatically dissociated atria

and ventricles were reacted separately with Alde�uor, andALDHbr cells were sorted by FACS and placed in CorningCostar 6-well plates (Sigma)with no extracellularmatrix pro-tein coating. Ventricular ALDHbr cells grew poorly in culture,possibly as a result of a very long sorting procedure due tothe scarcity of ALDHbr cells in the ventricular population.erefore, atrial cells from 8-week-old mice were used insubsequent experiments (𝑛𝑛 𝑛 𝑛 per experiment), allowingfor a marked abbreviation of the sorting procedure. ALDHbr

cells were cultured in MesenCult medium (MesenCult MSCBasal Medium supplemented with serum-containing Mes-enCult MSC Stimulatory Supplements-Mouse; STEMCELLTechnologies). ALDHdim cells were studied for comparison.To assess the impact of the culture medium on cell pheno-type, ALDHbr sorted cells were also cultured in RPMI/FCSmedium (RPMI 1640; Gibco, supplemented with 10% fetalcalf serum). e AlamarBlue assay (Promega, Madison, WI,USA) was used to assess cell viability and growth. In a sepa-rate experiment, cells were cultured in the presence of ima-tinib (methanesulfonate salt, 0.1–10 𝜇𝜇M; LC Laboratories,Woburn, MA, USA), a inhibitor of receptor tyrosine kinasesencoded by c-kit and platelet-derived growth factor receptorB (PDGFRB). Cells were incubated with imatinib for 42 h (intriplicates) before AlamarBlue was added. Absorbance wasmeasured at 570 nm and 595 nm every hour during 6 hours,and AlamarBlue reduction was calculated.

2.7. Differentiation Potential of Expanded 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑏𝑏𝑏𝑏 Cells.Cells sorted on the basis of high ALDH activity were culturedfor 8 months in MesenCult, which was then replaced by NHChondroDiff Medium (for 21 days), NH OsteoDiff Medium(7 days), or AdipoDiff Medium (14 days; all from Miltenyi).Chondrogenic differentiation was assessed by immunostain-ing for type II collagen. To assess osteogenic and adipogenicdifferentiation, cells were incubated with 1% PFA for 10min,followed by 2% Alizarin Red solution for 5min and Oil Red-O solution for 15min, respectively, and by 3 PBS-washes.

3. Results

3.1. 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑏𝑏𝑏𝑏 Cells in Neonatal and Young Adult Hearts.Freshly isolated nonmyocytic cells from enzymatically dis-sociated, neonatal or 8-week-old hearts were analyzed by�ow cytometry using the Alde�uor reagent. ALDHbr cellsamounted to 1.00 ± 0.59% of all nonmyocytic cells in theneonatal heart (𝑛𝑛 𝑛 𝑛) and to 0.99 ± 0.55% of these cells in theyoung adult heart (𝑛𝑛 𝑛 𝑛, NS; Figure 1). Within the ALDHbr

population, the percentage of ALDHvery-br cells in neonatalhearts was higher than in young adults (52.𝑛1 ± 6.8𝑛% versus19.𝑛9 ± 2.𝑛2%; 𝑃𝑃 𝑃 0.05).

3.2. 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑏𝑏𝑏𝑏 Cells in Atria and Ventricles. In young adultmice, ALDHbr cells in the atrial population were morefrequent than in the ventricular (10.29 ± 8.44% versus 1.05 ±0.78%; 𝑛𝑛 𝑛 𝑛; 𝑃𝑃 𝑃 0.05; Figure 2). Atrial ALDHbr

cells predominantly exhibited SSClo properties but a subset


1000

800

600

400

200

0

10008006004002000

SS

C

FSC

DEAB

R1 R2

0.1%

R3

0%

1000

800

600

400

200

0

100 101 102 103 104

ALDH

R1 R2

R3

1000

800

600

400

200

0

100 101 102 103 104

ALDH

No DEAB

SS

C

SS

C 1.5%

0.7%

Neonatal

(a)

1000

800

600

400

200

0

10008006004002000

SS

C

FSC

1000

800

600

400

200

0

SS

C

1000

800

600

400

200

0

SS

C

100 101 102 103 104

ALDH

100 101 102 103 104

ALDH

R2

0.1%

R30%

R2

R3

0.83%

0.12%

DEAB No DEAB

8 week-old

(b)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

AL

DH

br/

very

-br

cell

s (%

)

Neonatal

ALDHbr

ALDHvery-br

8 week-old

(c)

F 1: Identi�cation of ALDHbr cells in populations isolated from whole hearts from neonatal and young adult mice. Flow cytometricanalysis of ALDH activity (a). Neonatal mice. Le panel: cells were selected according to forward scatter (FSC) and side scatter (SSC)properties using the gated region. Middle panel: cells incubated with Alde�uor substrate and the speci�c inhibitor of ALDH, DEAB wereused to establish baseline �uorescence of these cells and to de�ne the ALDHbr (R2) and the ALDHvery-br (R3) region as less than 0.1% and 0%of total events, respectively. Right panel: cell incubationwith Alde�uor substrate in the absence of inhibitor induced a shi in FL1 �uorescencede�ning the ALDHdim (R1), the ALDHbr (R2), and the ALDHvery-br (R3) populations (b). Eight-week-old mice (c). Bar histogram showingALDHbr and ALDHvery-br cell percentages (mean, SD) in neonatal and young adult mice. 𝑃𝑃 value refers to the difference between neonataland 8 week-old mice.


showed intermediate-to-high SSC properties. VentricularALDHbr cells exhibitedmore homogeneous SSClo properties.

3.3. Immunophenotype of Freshly Isolated𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑏𝑏𝑏𝑏 Cells. Weused a panel of mAbs to determine the cell-surface markerpro�le of ALDHbr, ALDHvery-br, and ALDHdim cells in thefreshly isolated, nonmyocytic cardiac-derived population in8 weeks old as well as in 24 months old mice (Figure3). Aging hearts contained 1.36% ALDHbr cells, of which36.8% were ALDHvery-br (pooled sample from 5 hearts).A majority of ALDHbr and ALDHvery-br cells from youngadult hearts stained positive for Sca-1 (≈60% and ≈70%,resp.). Substantial subsets (≈25–50%) of ALDHbr cells stainedpositive for CD34, CD106, CD44, CD90, and CD105 (Figure4). e ALDHbr population was signi�cantly enriched forcells expressing Sca-1, CD90, CD34, CD106, andCD44 (fold-changes: 4.2, 4.6, 6.2, 24.1, and 34.3, resp.,) whereas it wassigni�cantly depleted for cells expressing CD45 (leukocytecommon antigen), CD31 (an endothelial marker), and CD38(fold-changes: 0.5, 0.36, and 0.22 fold). e immunopheno-type of ALDHbr cells derived from aging hearts was similarto that of ALDHbr cells from young adult hearts.

3.4. mRNA Expression of Selected ALDH Isoforms. mRNAexpression of ALDH1A1, ALDH1A2, ALDH1A3, andALDH2 isoforms was measured by real-time RT-PCR inpuri�ed ALDHdim, ALDHbr, and ALDHvery-br cells (Figure5). Gene transcripts of all of the four ALDH isoforms couldbe detected in ALDHbr and ALDHvery-br cells, whereas onlyALDH1A1 and ALDH1A2 were detectable in ALDHdim

cells. In ALDHbr and ALDHvery-br cells, ALDH1A1 wasexpressed at higher levels than ALDH1A3 (2−ΔCt = 4.0 and4.7, resp.,) and ALDH2 (2−ΔCt = 4.8 and 8.8, resp.,) whileALDH1A2 was expressed at higher levels than ALDH1A3(2−ΔCt = 1.7 and 12.3, resp.,) and ALDH2 (2−ΔCt = 2.1and 23.2, resp.). In addition, ALDH1A1 was expressed athigher levels in ALDHbr and ALDHvery-br cells comparedto ALDHdim cells (2−ΔΔCt = 2.0 and 3.5, resp.). ALDH1A2expression was higher in ALDHvery-br cells compared toALDHdim cells (2−ΔΔCt = 4.9) and was lowest in ALDHbr cells(2−ΔΔCt = 0.5 versus ALDHdim). mRNA expression of threegenes implicated in angiogenesis was also measured. 2−ΔCt

values for ALDHdim, ALDHbr, and ALDHvery-br cells were asfollows: endoglin (CD105): 11.70, 10.41, and 14.86; ephrinB4: 2.98, 0.90, and 2.60; angiopoietin 1: undetectable, 0.01,and 0.008, respectively.

3.5. Ex Vivo Culture-Expansion of 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑏𝑏𝑏𝑏 Cells. ALDHbr

cells were puri�ed from the atrial population, which wasenriched for these cells compared to ventricles, in orderto abbreviate the FACS procedure, thereby limiting celldamage. Puri�ed atrial ALDHbr cells grew in culture, whereasALDHdim cells did not, even when plated at a 10-fold higher

density compared to their ALDHbr counterparts (Figure6(a)). However, >95%ALDHdim cells appeared to be viable,as assessed by DAPI, immediately aer the FACS procedure.Atrially derived bulk populations of nonmyocytic cells grewpoorly. Puri�ed ALDHbr cells gave rise to small numbers ofcell colonies, each one apparently originating from a singlecell, which formed a monolayer of plastic-adherent cells,which could be expanded for more than 40 passages (laterpassages were not tested). In one cell culture, the percentageof ALDHbr cells at passage 7 was 11%; however, this parame-ter was not systematically measured at various time points.Growth rates in MesenCult medium were higher than inRPMI/FCS medium (Figure 6(b)). is was con�rmed usingthe AlamarBlue assay (Figure 6(c)). Imatinib inhibited cellgrowth in a dose-dependent manner (Figure 6(d)).

3.6. Immunophenotype of Culture-Expanded 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑏𝑏𝑏𝑏 Cells.e marker pro�le of expanded ALDHbr cells (P11-13) wasanalyzed by �ow cytometry (Figure 7). Cells expanded inMesenCult medium stained positive for Sca-1, CD29, CD44,CD105, CD106, and, in part, CD146 and CD14.ey stainednegative for CD45, CD11b, CD31, and CD133. To assesswhether the culture medium affected marker expression,cells were also grown in RPMI/FCS. e immunophenotypeof these cells was similar to those grown in MesenCult,although larger cell subsets stained positive for c-kit, CD140b(PDGFRB) and NG2 chondroitin sulfate proteoglycan (Fig-ures 7(b) and 7(c)). NG2 essentially colocalized with CD140band CD146 (Figure 7(d)). c-kit expression was demonstratedby immunocytochemistry (Figure 7(e)).

3.7. Expanded Cells Express Sarcomeric 𝛼𝛼-Actinin and Dif-ferentiate along Mesenchymal Lineages. Culture-expandedALDHbr cells could be induced to differentiate along adi-pogenic, osteogenic, and chondrogenic lineages in appropri-ate culture media, as evidence by staining with Oil red-O,Alizarin red, Alcian blue (not shown), and type II collagenimmunostaining, respectively (Figures 8(a), 8(c) and 8(g)).Cells cultured in MesenCult stained positive for sarcomeric𝛼𝛼-actinin (Figure 8(e)) but negative for 𝛼𝛼-SMA and vonWillebrand factor (not shown).

4. Discussion

Growing evidence suggests high ALDH activity may be acommon feature shared by stem and progenitor cells acrossnormal tissues, as well as in cancer. Human UCB and BMcells possessing high ALDH activity have shown angiogenicactivity in preclinical studies [30–32] and have been usedsafely in phase I clinical trials in patients with ischemiccardiovascular disease [33–35].

We characterized, for the �rst time, ALDHbr cells isolatedfrom the heart. Approximately one percent of all nonmy-ocytic cells present in the young adult mouse heart wereALDHbr, as measured by Alde�uor stain (Figure 1). eneonatal heart contained a similar number of ALDHbr cells,although with a higher proportion of cells exhibiting very


1000

800

600

400

200

0

10008006004002000

SS

C

FSC

1000

800

600

400

200

0

SS

C

100 101 102 103 104

ALDH

1000

800

600

400

200

0

SS

C

100 101 102 103 104

ALDH

1000

800

600

400

200

0

10008006004002000

SS

C

FSC

1000

800

600

400

200

0

SS

C

100 101 102 103 104

ALDH

1000

800

600

400

200

0

SS

C

100 101 102 103 104

ALDH

Atrial cells DEAB No DEAB

DEAB No DEABVentricular cells

R1 R2

0.1%

0.1%1.3%

15.03%

R1 R2

R1 R2 R1 R2

(a)

0

5

20

AL

DH

br

cell

s (%

)

15

10

AtrialVentricular

(b)

F 2: Identi�cation of ALDHbr cells in the atrial and ventricular population from 8 week-old mice. Flow cytometric analysis of ALDHactivity (a). Le panels: atrial and ventricular cells were selected according to forward scatter (FSC) and side scatter (SSC) properties usingthe gated region. Middle panels: cells incubated with Alde�uor substrate and the speci�c inhibitor of ALDH, D�AB, were used to establishbaseline �uorescence of these cells and to de�ne the ALDHbr region (R2) as less than 0.1% of total events. Right panels: cell incubation withAlde�uor in the absence of inhibitor induced a shi in FL1 �uorescence de�ning the ALDHdim (R1) and the ALDHbr (R2) population. Inthe example shown, not all of the atrial cells exhibiting a shi in FL1 �uorescence in the absence of inhibitor were found in the gated regionR2, suggesting that atrial ALDHbr cells may have been underestimated (b). Bar histogram showing percent ALDHbr cells (mean, SD) in theventricular and atrial population.

high ALDH activity (ALDHvery-br) compared to the youngadult heart. e frequency of ALDHbr cells in the populationisolated from the atria was approximately 10-fold higher thanin cells derived from ventricles (Figure 2). is observationhas analogies with previous reports on increased numbers of

putative stem/progenitor cells, such as DNA-label retainingcells in rodents [36, 37] and c-kit+ cells in human [38], inatria relative to ventricles. In the present study, ALDHbr

cells isolated from atria and ventricles exhibited slightlydifferent side scatter (SSC) properties. Most atrial ALDHbr


1000

800

600

400

200

0

10008006004002000

SS

C

FSC ALDH

DEAB

R1 R2

0.1%

R30%

No DEAB

R1 R2

R3

0.83%

0.31%

100 101 102 103 104100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104100 101 102 103 104

100 101 102 103 104100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104100 101 102 103 104

100 101 102 103 104

100 101 102 103 104 100 101 102 103 104100 101 102 103 104

100 101 102 103 104 100 101 102 103 104100 101 102 103 104

100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

0

2

4

6

8

10

0

2

4

6

8

10

0

2

4

6

8

10

0

2

4

6

8

10

0

2

4

6

8

10

0

2

4

6

8

10

0

2

4

6

8

10

0

2

4

6

8

10

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

200

400

600

800

1000

0200

400

600

800

1000

0200

400

600

800

1000

0

200

400

600

800

1000

0

200

400

600

800

1000

0

200

400

600

800

1000

0

200

400

600

800

1000

0

200

400

600

800

1000

0

200

400

600

800

1000

0

200

400

600

800

1000

0

200

400

600

800

1000

0

200

400

600

800

1000

0

200

400

600

800

1000

0

200

400

600

800

1000

0

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

CD31

CD49a

CD105 CD106

CD90

CD40

200

400

600

800

1000

0

100 101 102 103 104

CTRL Sca-1 CD44

M1 M1 M1

M1M1

M1 M1

M1

M1

M1

M1

M1

M1

M1M1

M1

M1

M1

M1

M1

M1

M1

M1

M1

M1

M1

M1 M1 M1 M1

M1 M1 M1 M1

M1

M1

M1

M1 M1 M1 M1

CTRL CD34

CTRL Lin CD38

0.2% 15.5%44.5% 0.5% 0% 0.3% 10.5% 3% 6.8% 2%

58.5% 29.7% 29% 0% 2.1% 1.9% 2.9% 27.3%

76.2% 8.4% 11.6% 0% 0.5% 1.1% 0.4% 13.6%

13.5%1.6%7.4%11.2%0.2%

13.1% 32.8% 4.5%

13.6%0.4%0.5%

1.1%

1.1%

8 week-old

1000

800

600

400

200

0

SS

C

100 101 102 103 104

1000

800

600

400

200

0

SS

C

ALDH

100 101 102 103 104

CD45

ALDHdim

ALDHbr

ALDHvery-br

(a)

1000

800

600

400

200

0

10008006004002000

SS

C

FSC

DEAB

R1 R2

0.1%

R30%

No DEAB

R1 R2

R3

1.6%

0.3%

200400600800

1000

0

100 101 102 103 104

200400600800

1000

0

100 101 102 103 104

100 101 102 103 104

200400600800

1000

0

100 101 102 103 104

200400600800

1000

0

100 101 102 103 104

200400600800

1000

0

100 101 102 103 104

200400600800

1000

0

100 101 102 103 104

200400600800

1000

0

100 101 102 103 104

200400600800

1000

0

100 101 102 103 104

200400600800

1000

0

100 101 102 103 104

200400600800

1000

0

200400600800

1000

0200400600800

1000

0200400600800

1000

200400600800

1000

00200400600800

1000

0

100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

0

5

10

15

20

100 101 102 103 104

0

5

10

15

20

100 101 102 103 104

0

5

10

15

20

100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

100 101 102 103 104 100 101 102 103 104 100 101 102 103 104100 101 102 103 104

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

1520

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

0

5

10

15

20

CTRL Sca-1 CD44

0.2%

CD31 CD105 CD106CTRL CD34

CTRL Lin CD38

Flk1

c-Kit CD133

26.6% 25.9% 3.9% 1.5% 32.6% 2.5% 17.4% 4.1%

64.3% 26.2% 34.9% 0.5% 7.3% 2% 10.2% 35.5%

83.6% 5.1% 12.7% 0.2% 6.1% 1.2% 7.6% 4.8%

39.2%1.8%1.8% 1.4%

8.9%3%0.4% 0.4%

0%0% 1.7% 7.3%

M1 M1 M1 M1 M1

M1 M1 M1 M1

M1 M1 M1 M1

M1M1 M1 M1 M1

M1 M1 M1 M1

M1 M1 M1 M1

M1 M1 M1 M1

M1 M1 M1 M1

M1 M1 M1 M1

1000

800

600

400

200

0

SS

C

ALDH

100 101 102 103 104

ALDH

100 101 102 103 104

24 month-old

1000

800

600

400

200

0

SS

C

0.2%

CD45

M1

ALDHdim

ALDHbr

ALDHvery-br

(b)

F 3: Freshly isolated ALDHbr cells from young adult hearts (upper panel; (a)) and from aging hearts (lower panel; (b)) were analyzedby �ow cytometry for ALDH expression in combination with the expression of cell-surface markers. Cells were incubated with Alde�uorsubstrate and an inhibitor of ALDH (D�AB), or with Alde�uor alone, to establish R�, R�, and R3 gates for ALDHdim (blue), ALDHbr, andALDHvery-br cells (both green), respectively. ese populations were subsequently analyzed for expression of markers. Cytograms for eachsurfacemarker in each cell population are shown (the surfacemarkers indicated for the cytograms ofALDHdim cells also refer to the cytogramsof the respective ALDHbr and ALDHvery-br populations; corresponding cytograms are aligned vertically. Percentages of positive cells areindicated (CTRL; isotype-matched irrelevant IgG).


0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100DepletedNeither enriched nor depleted

Sca

-1

CD

34

CD

90

CD

106

CD

44

Enriched in ALDHbr cells

Po

siti

ve c

ells

(%

)

Po

siti

ve c

ells

(%

)

Po

siti

ve c

ells

(%

)

Flk

-1

CD

45

CD

31

CD

38

CD

105

CD

133

c-K

it

MH

C-I

I

CD

40 Lin

All cells

ALDHbr

ALDHvery-br

2x

1.8x

2.9x

4.5x

4.2x

6.2x4.6x

24.1x

34.3x

All cells

ALDHbr

ALDHvery-br

All cells

ALDHbr

ALDHvery-br

∗∗∗

∗

∗∗

∗

∗

∗∗

∗∗∗

∗∗∗

∗∗

∗∗

∗

∗

F 4: Flow cytometric analysis of the expression of cell-surface antigen markers in freshly isolated nonmyocytic cardiac cells from8-week-old mice. Data are percentages (mean, SD) of positive cells for the indicated markers, subdivided into three groups: thosesigni�cantly enriched, those signi�cantly depleted, and those neither enriched nor depleted in the ALDHbr population. Yellow barsindicate bulk populations of cardiac-derived cells, green bars indicate ALDHbr cells, and blue bars indicate ALDHvery-br cells. Numbers areenrichment/depletion factors for the indicated surface markers in the ALDHbr population (∗∗∗𝑃𝑃 𝑃 𝑃𝑃𝑃𝑃𝑃, ∗𝑃𝑃 𝑃 𝑃𝑃𝑃𝑃 versus ALDHdim cells).

cells showed SSClo properties but a small subset possessedintermediate SSC properties. In contrast, ventricular ALDHbr

cells uniformly exhibited SSClo properties.e original char-acterizations of ALDHbr cells from humanUCB revealed thatthe ALDHbrSSClo population was enriched 50- to 100-foldfor primitive hematopoietic progenitors [11, 15]. In severalstudies, SSClo properties therefore were associated withthe traditional de�nition of ALDHbr hematopoietic cells.However, other studies focused on ALDHbrLin− (lineage-depleted) cells [13, 14, 16]. �e de�ned ALDHbr cells basedsolely on ALDH activity, regardless of SSC and lineage prop-erties. is methodological approach affected ALDHbr cellcounts onlymarginally, given the predominantALDHbrSSClo

phenotype. Nevertheless, the slightly different SSC propertiesof atrial and ventricular cells may suggest differences in thecellular composition of the ALDHbr population dependingon the cardiac chamber of origin. is issue remains to beaddressed in future studies.

Freshly isolated ALDHbr cells exhibited a heterogeneousphenotype, re�ecting amixture of different cell types (Figures3 and 4). is population was signi�cantly enriched for cellsexpressing Sca-1, the early hematopoietic and EPC markerCD34, the MSC marker CD90, the hyaluronic acid receptorCD44, and vascular cell adhesionmolecule-1 (CD106). CD44and CD106 were the most highly enriched markers inthe ALDHbr versus ALDHdim population (34.3 and 24.1fold-increases, resp.). Previous studies associated CD44 and

CD106 expression with MSCs [39]. However, a recent studyshowed that freshly isolated murine and human BM stromalcells did not express CD44 andCD106, but they could acquireexpression of these antigens in vitro [40]. CD44+CD24−cells have been proposed to represent cancer stem cells inbreast cancer [41], as well as in head and neck squamouscell carcinoma (the latter includes an ALDH1+CD44+CD24−

stem cell subset) [42]. In the present study, ALDHvery-br

cells were highly enriched for Sca-1 and CD90, but onlymoderately so for CD34, possibly re�ecting a predominantMSC-like subpopulation with fewer hematopoietic/vascularprogenitors compared to ALDHbr cells with lower levelsof ALDH activity. e ALDHbr population was signi�-cantly depleted for common leukocyte antigen (CD45), theendothelial marker CD31, and CD38 (an antigen expressedby the differentiating progeny of CD34+CD38– hematopoi-etic/vascular progenitors). ALDHvery-br cells were furtherdepleted for lineage differentiation markers compared toALDHbr cells. us, the ALDHbr population was enrichedfor MSC-like progenitors and depleted for differentiatinghematopoietic and endothelial cells.

Puri�ed ALDHbr atrial cells could be expanded ex vivo,whereas their ALDHdim counterparts could not. Becausethe atrial population contained more ALDHbr cells thanthe ventricular population, atrial cells were used for thepuri�cation of ALDHbr cells in order to shorten the sortingprocedure, and therefore to attenuate cell damage. Puri�ed


256

192

128

64

0

SS

C

FSC

256192128640

256

205

154

102

51

0

100 101 102 103 104 105

SS

C

256

192

128

64

0

SS

C

FSC

256192128640

256

205

154

102

51

0

100 101 102 103 104 105

SS

C

With DEAB No DEAB No DEAB

ALDH ALDH

R1R2 R3

R1R2 R3

With DEAB

(a)

0.1

1

10

100

ALDH1A1 ALDH1A2 ALDH1A3 ALDH2

ND ND

ALDHbr

ALDHvery-br

ALDHdim

mR

NA

exp

ress

ion

(2

−∆

Ct )

(b)

F 5: Real time RT-PCR analysis of gene transcripts of selected ALDH isoforms (a). FAC sorting of freshly isolated ALDHdim (R1),ALDHbr (R2), and ALDHvery-br (R3) atrial cells from 8-week-old mice (b). RT-PCR analysis of ALDH1A1, ALDH1A2, ALDH1A3, andALDH2 gene transcripts (data are 2−ΔCt values using GUSB as a reference gene). ND: not detectable.

ALDHbr cells formed a few plastic-adherent colonies ineach culture well. Each colony apparently originated froma single cell, providing observational evidence of clonalexpansion, even though this was not �rmly establishedusing a clonogenic assay. e mechanism responsible forthe selective growth of ALDHbr cells in culture remainsunclear. To initially address this question, we determinedthe respective contributions of selected ALDH isozymesto ALDH activity, measured by Alde�uor stain. Fi�eenALDH isozymes have been described in mice [43] and 19in human [9, 44]. ey play different biological roles thatvary among cell types and species. ALDH1A1, the ALDHisozyme most highly overexpressed in human BM ALDHbr

versus ALDHdim cells [44], was believed to be responsible forALDHactivity, measured byAlde�uor stain, in BM and othertissues.However, recent �ndings in geneticallymodi�edmicedemonstrated that ALDH1A1 was dispensable for stem cellfunction in the mouse hematopoietic and nervous systems[45]. ALDH1 was identi�ed as a marker of normal andmalignant human mammary stem cells and a predictor ofpoor clinical outcome in breast cancer [28]. In a distinctstudy, shRNA knock-down data indicated that ALDH1A3,not ALDH1A1, correlated best with ALDH activity in breast

cancer stem cells [46]. ALDH1A1 andALDH3A1were highlyexpressed in nonsmall cell lung cancer, and knock-down ofthese ALDH isoformswere associatedwith in vitro functionalchanges in the proliferation and motility of these cells [47].Human ALDH1A1, ALDH1A2, ALDH1A3, and ALDH8A1function in retinoic acid cell signaling via retinoic acidproduction by oxidation of all-trans-retinal and 9-cis-retinal[6, 46]. ALDH2, a mitochondrial isozyme that mediates boththe detoxi�cation of reactive aldehydes and the bioactivationof nitroglycerin to nitric oxide, mediates cytoprotection inthe heart [48]. Recent data indicated that the aldehyde-oxidizing activity of mouse HSCs, measured by Alde�uorstain, was due to the ALDH2 isozyme and correlated withprotection against acetaldehyde toxicity [49]. Mice de�cientin both Fanconi anemia pathway-mediated DNA repair andacetaldehyde detoxi�cation showed more than a 6��-foldreduction in the HSC pool.

Based on these reports, we measured the expressionof ALDH1A1, ALDH1A2, ALDH1A3, and ALDH2 genetranscripts in puri�ed ALDHdim, ALDHbr, and ALDHvery-br

cardiac-derived cells (Figure 5). Expression of ALDH1A1was lowest in ALDHdim cells, intermediate in ALDHbr

cells, and highest in ALDHvery-br cells (2−ΔΔCt = 3.5 using


ALDHbr (d7) ALDHbr (d12) ALDHdim (d7)

(a)

Passage

0

5

10

15

20

25

30

35

0 50 100 150 200

RPMI/FCS

MesenCult

Days

P0

P7

P11

P29

(b)

Hours

0

5

10

15

20

25

1 2 3 4 5 6

RPMI/FCS

MesenCult

AR

(%

)

(c)

0

20

40

60

80

100

0 0.1 1 10

Rel

ativ

e gr

ow

th (

%)

(d)

F 6: Ex vivo culture-expansion of ALDHbr atrial cells (a). Photomicrographs of ALDHbr cells cultured for 7 and 12 days in MesenCultmedium. ALDHdim cells were dead by day 7 (b). Growth curves of cells in different growth media. Cells were cultured in MesenCult mediumup to P10 (arrow). From P11 on, MesenCult was either maintained or replaced by RPMI/FCS. Symbols (squares and diamonds) indicate cellpassage (no data available from P1 to P7; dotted line) (c). AlamarBlue assay on cells (P40) cultured in either MesenCult or RPMI/FCS (AR%:percent AlamarBlue reduction) (d). Dose-response study of Imatinib with respect to growth inhibition of expanded ALDHbr cells (P25).

ALDHdim as a reference population). By contrast, expressionof ALDH1A2 was lowest in ALDHbr cells, intermediate inALDHdim cells, and highest in ALDHvery-br cells (2−ΔΔCt =4.9 versus ALDHdim). Expression of ALDH1A3 and ALDH2was detectable in ALDHbr and ALDHvery-br cells, albeit atlower levels compared to ALDH1A1 and ALDH1A2, butnot in ALDHdim cells. us, ALDH1A1 correlated best withALDH activity; however, contributory roles of ALDH1A2,ALDH1A3, and ALDH2 to ALDH activity are possible basedon our data. While the present analysis included the fourALDH isoformsmost oen associated with ALDH activity in

previous reports [44–49], it should be completed to includethe remaining ALDH isoforms. In addition, shRNA knock-down experiments are needed for identifying the ALDHisozyme(s) responsible for ALDH activity in these cells.

Recent data suggested BM and UCB ALDHbr cellsmight repair ischemic tissues in vivo by releasing angio-genic factors [30–32]. A comparison of gene expressionpro�les by ALDHbr and ALDHdim human BM cells identi�edthree angiogenic factors (endoglin/CD105, ephrin B4, andangiopoietin-1) as the most highly overexpressed factors inALDHbr versus ALDHdim cells (fold-changes: 66.9, 64.6, and


200

160

120

80

40

0

100 101 102 103 104

200

400

600

800

1000

SS

C

FSC

0

200

160

120

80

40

0

200 400 600 80010000

ALDHbr cells in MesenCult medium

CTRL-APC

CTRL-APC

CTRL-PE CTRL-FITC

Sca-1 CD31 c-Kit CD44

CD11b CD14 CD140b CD133

CD105 CD106 CD29 CD146

CD90200

160

120

80

40

0

200

160

120

80

40

0

200

160

120

80

40

0

200

160

120

80

40

0

200

160

120

80

40

0

200

160

120

80

40

0

200

160

120

80

40

0

200

160

120

80

40

0

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

0

20

40

60

80

100

120

0

20

40

60

80

100

0

20

40

60

80

100

250

200

150

100

50

00

20

40

60

80

100

120

0

20

40

60

80

100

120

0

20

40

60

80

100

120

0

20

40

60

80

100

120

M1

M1

M1

M2

M1

M1

M1

M1

M1

M2

M1M1

M1

M2

M1

M1

M1

M2

M1

M1

M1

CD45

(a)

200

400

600

800

1000

SS

C

FSC

0

200 400 600 80010000

200

160

120

80

40

0

100 101 102 103 104

200

160

120

80

40

0

100 101 102 103 104

200

160

120

80

40

0

100 101 102 103 104

200

160

120

80

40

0

100 101 102 103 104

100 101 102 103 104

200

160

120

80

40

0

200

160

120

80

40

0

200

160

120

80

40

0

200

160

120

80

40

0

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

100 101 102 103 104

200

160

120

80

40

0

200

160

120

80

40

0

100 101 102 103 104

100 101 102 103 104

100 101 103 104102

100 101 102 103 104

100 101 102 103 104

00 101 102 103 104

200

160

120

80

40

0

200

160

120

80

40

0

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

120

250

200

150

100

50

0

CTRL-APC

CTRL-PE CTRL-FITC

Sca-1 CD31 c-Kit CD44

CD45 CD11b CD14 CD140b CD133

CD105 CD106 CD146

CD90

ALDHbr cells in RPMI/FCS medium

M2

M1

M1 M1

M2

M1

M2

M1

M1

M2

M2

M1

M1

M2

M1

M2

M1

M2

M1

(b)

0

20

40

60

80

100

Adhesion moleculesMSC markers

Hematopoetic markers

Pericyte markers

Po

siti

ve c

ells

(%

)

CD

31

c-K

it

CD

14

CD

140b

CD

133

CD

105

CD

106

CD

146

CD

90

Sca

-1

CD

44

CD

29

NG

2

CD

34

CD

11b

Stem cell-associated markers

CD

45

(c)

100 101 102 103 104100101102103104

100 101 102 103 104100101102103104

100 101 102 103 104100101102103104

CTRL-PE

NG2-PECD146-FITC

CTRL-FITC FL1-height

CTRL-A

PC

FL4-height

FL4-height

100 101 102 103 104100101102103104

100 101 102 103 104100101102103104

CD140b-A

PC

100 101 102 103 104100101102103104

CD140b-A

PC

CD146-FITC

NG2-PE

1

12

10

56

22

59

335

35 74

20

(d)

F 7: Continued.


c-Kit

Hoechst 33342

Hoechst 33342 Merge

MergeCTRL

(e)

F 7: (a)/(b) Flow cytometric analysis of ALDHbr atrial cells expanded in MesenCult (P11) or RPMI/FCS medium (P13). (c)Percentages of cells (mean, SD) expressing the indicated markers (data from 3 separate experiments in MesenCult). (d) Two-color analysisof pericyte/perivascular markers (NG2, CD146, and CD140b; numbers are cell percentages). (e) c-kit immunostaining of expanded ALDHbr

cells. Nuclear staining with Hoechst 33342 (CTRL, secondary Ab only as a control).

10x

(a)

40x

(b)

20x

(c)

20x

(d)

20x

(e)

20x

(f)

10x

(g)

F 8: Differentiation potential of expanded ALDHbr cells. (a) Photomicrograph showing intracellular fat droplets in cells cultured inadipogenic medium. (b) Positive Oil Red-O staining of intracellular droplets. (c) Positive type-II collagen immunostaining (green) in cellscultured in chondrogenic medium. (d) Type-II collagen immunostaining of leg cartilage, as a positive control. (e) Positive sarcomeric 𝛼𝛼-actinin immunostaining (green) of an ALDHbr cell (the protein does not show clear sarcomeric distribution). (f) Corresponding negativecontrol (secondary Ab only). (g) Positive Alizarin Red staining of expanded ALDHbr cells cultured in osteogenic medium.

7.0, resp.) [44]. We therefore measured mRNA expressionof these angiogenic genes in cardiac-derived cells but foundno correlation with ALDH activity (besides the preliminaryobservation that angiopoietin-1 transcripts were detectablein ALDHbr and ALDHvery-br cells but not in ALDHdim cells).e angiogenic potential of cardiac-derived ALDHbr cellsremains to be addressed in future studies. Previous reportsshowed that ALDHbr MSCs from human UCB were more

responsive to hypoxia than theirALDHdim counterparts, withupregulation of Flt-1, CXCR4, and angiopoietin-2 [50]. esame group reported that ALDHdim EPCs, but not ALDHbr

EPCs, from human UCB upregulated hypoxia-inducible fac-tor proteins as well as VEGF, CXCR4, and GLUT-1 mRNAsunder hypoxic conditions [22].e introduction ofALDHdim

EPCs signi�cantly reduced ischemic tissue damage in amouse �ap model, whereas ALDHbr EPCs were ineffective.


ese �ndings suggested varying angiogenic activities ofALDHbr cells depending on the cell type studied.

In the present study, plastic-adherent cell clusters grewrapidly and could be passaged more than 40 times (laterpassages were not tested). Growth rates in complete Mes-enCult medium were higher than in RPMI/FCS (Figure6). Imatinib, which selectively inhibits PDGFRB (CD140b)and c-kit tyrosine kinases with a similar potency [51],reduced cell proliferation in a dose-dependent manner.Expression of PDGFRB and c-kit was detected in subsetsof expanded cells. e effect of Imatinib might re�ect arole for PDGFRB and/or c-kit in the proliferation of thesecells. Previous studies showed that ALDHbr MSCs fromUCBproliferated more than their ALDHdim counterparts [50],whereas ALDHdim EPCs proliferated more than ALDHbr

EPCs [22]. Proliferative properties correlatedwith angiogenicactivities in these populations. Again, the association ofALDH and proliferation may vary among cell types. eimmunophenotype of expandedALDHbr cardiac cells resem-bled that of freshly isolated ALDHbr cells (Figure 7). Positivestains for pericytic markers (NG2, CD146, and CD140b)were consistent with a MSC-like phenotype. It has beenshown that perivascular cells of arterioles and capillariesfrommany tissues express pericytic markers and showMSC-like and angiogenic features [52–54]. In the present study,expanded cells stained positive for sarcomeric 𝛼𝛼-actinin,a cardiac marker, and exhibited differentiation potentialalong multiple mesenchymal lineages. Previous data showedthat aldehyde dehydrogenase activity did not increase thechondrogenic potential of human adipose-derived adultstem cells [55]. Recent reports showed that human BMcontained ALDHbr multipotential mesenchymal progenitorcells (MPCs), besides ALDHbr HSCs and EPCs [12, 56].eseMPCs strongly expressedALDH, expressed embryonicmarkers not present in mesenchymal cells, and could bedifferentiated to microvascular endothelial cells and mes-enchymal cells. e latter lacked strong ALDH expressionand could not be induced to back differentiate into MPCs.

We observed a loss of ALDHbr cells with cell passage(e.g., only 11% of cells at passage 7 were still ALDHbr).Whether this loss of ALDHbr cells with time re�ected cellularadaptation to ex vivo conditions or the replacement of earlyprogenitors by a progeny of late progenitor and precursorcells remains unclear. e observation that growth rates andthe immunophenotype remained stable for more than 30 and10 passages, respectively, seemed not to support cell differen-tiation. ALDH as a stem/progenitor cell marker has chie�ybeen used in freshly isolated cells and tissues, as opposedto expanded cell populations. e signi�cance of ALDHin culture-expanded cells is unclear, as is its contributionto stem cell fate. e aforementioned observation that themesenchymal cell progeny ofALDHbr humanBMcells lackedstrong ALDH expression and could not be induced to backdifferentiate into MPCs suggested a parallel loss of ALDHexpression and cell potency [56].

In conclusion, cardiac-derived ALDHbr cells are pro-genitor cells with MSC-like phenotype and function and

with superior ex vivo growth characteristics compared totheir ALDHdim counterparts. Further studies are needed forassessing the regenerative potential of these cells in animalmodels of heart disease.

Acknowledgments

is work is supported by the Swiss Heart Foundation(Berne, Switzerland), the Fondation Vaudoise de Cardiologie(Lausanne, Switzerland), the Cecilia Augusta Foundation(Lugano, Switzerland), the METIS Foundation Sergio Man-tegazza (Lugano, Switzerland), and the “Fondazione per laricerca sulla trasfusione e sui trapianti” (Lugano, Switzer-land).

References

[1] H. Oh, S. B. Bradfute, T. D. Gallardo et al., “Cardiac progenitorcells from adult myocardium: homing, differentiation, andfusion aer infarction,” Proceedings of the National Academy ofSciences of the United States of America, vol. 100, no. 21, pp.12313–12318, 2003.

[2] A. P. Beltrami, L. Barlucchi, D. Torella et al., “Adult cardiac stemcells are multipotent and support myocardial regeneration,”Cell, vol. 114, no. 6, pp. 763–776, 2003.

[3] K. L. Laugwitz, A. Moretti, J. Lam et al., “Postnatal isl1+cardioblasts enter fully differentiated cardiomyocyte lineages,”Nature, vol. 433, no. 7026, pp. 647–653, 2005.

[4] K. L. Laugwitz, A. Moretti, J. Lam et al., “Erratum: Postnatalisl1+ cardioblasts enter fully differentiated cardiomyocyte lin-eages,” Nature, vol. 446, no. 7138, p. 934, 2007.

[5] J. Cai, M. L. Weiss, and M. S. Rao, “In search of “stemness”,”Experimental Hematology, vol. 32, no. 7, pp. 585–598, 2004.

[6] A. E. Balber, “Concise review: aldehyde dehydrogenase brightstem and progenitor cell populations from normal tissues:characteristics, activities, and emerging uses in regenerativemedicine,” Stem Cells, vol. 29, no. 4, pp. 570–575, 2011.

[7] D. J. Laird, A.W.De Tomaso, and I. L.Weissman, “Stem cells areunits of natural selection in a colonial ascidian,” Cell, vol. 123,no. 7, pp. 1351–1360, 2005.

[8] D. J. Laird, A. W. De Tomaso, and I. L. Weissman, “Erratum:Stem cells are units of natural selection in a colonial ascidian,”Cell, vol. 124, no. 3, pp. 647–648, 2006.

[9] S. A. Marchitti, C. Brocker, D. Stagos, and V. Vasiliou, “Non-P450 aldehyde oxidizing enzymes: the aldehyde dehydrogenasesuperfamily,” Expert Opinion on Drug Metabolism and Toxicol-ogy, vol. 4, no. 6, pp. 697–720, 2008.

[10] M. B. Kastan, E. Schlaffer, J. E. Russo, O. M. Colvin, C.I. Civin, and J. Hilton, “Direct demonstration of elevatedaldehyde dehydrogenase in human hematopoietic progenitorcells,” Blood, vol. 75, no. 10, pp. 1947–1950, 1990.

[11] R. W. Storms, A. P. Trujillo, J. B. Springer et al., “Isolationof primitive human hematopoietic progenitors on the basis ofaldehyde dehydrogenase activity,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 96, no.16, pp. 9118–9123, 1999.

[12] T. Gentry, S. Foster, L. Winstead, E. Deibert, M. Fiordalisi, andA. Balber, “Simultaneous isolation of human BM hematopoi-etic, endothelial and mesenchymal progenitor cells by �owsorting based on aldehyde dehydrogenase activity: implicationsfor cell therapy,” Cytotherapy, vol. 9, no. 3, pp. 259–274, 2007.


[13] J. P. Chute, G. G. Muramoto, J. Whitesides et al., “Inhibitionof aldehyde dehydrogenase and retinoid signaling induces theexpansion of human hematopoietic stem cells,” Proceedings ofthe National Academy of Sciences of the United States of America,vol. 103, no. 31, pp. 11707–11712, 2006.

[14] D. A. Hess, T. E. Meyerrose, L. Wirthlin et al., “Functional char-acterization of highly puri�ed human hematopoietic repopulat-ing cells isolated according to aldehyde dehydrogenase activity,”Blood, vol. 104, no. 6, pp. 1648–1655, 2004.

[15] R. W. Storms, P. D. Green, K. M. Safford et al., “Distincthematopoietic progenitor compartments are delineated by theexpression of aldehyde dehydrogenase and CD34,” Blood, vol.106, no. 1, pp. 95–102, 2005.

[16] D. A. Hess, L. Wirthlin, T. P. Cra et al., “Selection based onCD133 and high aldehyde dehydrogenase activity isolates long-term reconstituting human hematopoietic stem cells,” Blood,vol. 107, no. 5, pp. 2162–2169, 2006.

[17] D. J. Pearce and D. Bonnet, “e combined use of Hoechstefflux ability and aldehyde dehydrogenase activity to identifymurine and human hematopoietic stem cells,” ExperimentalHematology, vol. 35, no. 9, pp. 1437–1446, 2007.

[18] �. Christ, K. Lucke, S. Imren et al., “Improved puri�cationof hematopoietic stem cells based on their elevated aldehydedehydrogenase activity,” Haematologica, vol. 92, no. 9, pp.1165–1172, 2007.

[19] T. Gentry, E. Deibert, S. J. Foster, R. Haley, J. Kurtzberg, andA. E. Balber, “Isolation of early hematopoietic cells, includingmegakaryocyte progenitors, in the ALDH-bright cell popula-tion of cryopreserved, banked UC blood,” Cytotherapy, vol. 9,no. 6, pp. 569–576, 2007.

[20] P. Fallon, T. Gentry, A. E. Balber et al., “Mobilized peripheralblood SSCloALDHbr cells have the phenotypic and functionalproperties of primitive haematopoietic cells and their numbercorrelates with engrament following autologous transplanta-tion,” British Journal of Haematology, vol. 122, no. 1, pp. 99–108,2003.

[21] T. J. Povsic, K. L. Zavodni, F. L. Kelly et al., “Circulatingprogenitor cells can be reliably identi�ed on the basis of alde-hyde dehydrogenase activity,” Journal of the American Collegeof Cardiology, vol. 50, no. 23, pp. 2243–2248, 2007.

[22] M. Nagano, T. Yamashita, H. Hamada et al., “Identi�cation offunctional endothelial progenitor cells suitable for the treat-ment of ischemic tissue using human umbilical cord blood,”Blood, vol. 110, no. 1, pp. 151–160, 2007.

[23] S. Corti, F. Locatelli, D. Papadimitriou et al., “TransplantedALDHhiSSClo neural stem cells generate motor neurons anddelay disease progression of nmd mice, an animal model ofSMARD1,” Human Molecular Genetics, vol. 15, no. 2, pp.167–187, 2006.

[24] S. Corti, M. Nizzardo, M. Nardini et al., “Neural stem celltransplantation can ameliorate the phenotype of amousemodelof spinal muscular atrophy,” Journal of Clinical Investigation,vol. 118, no. 10, pp. 3316–3330, 2008.

[25] K. Vauchez, J. P. Marolleau, M. Schmid et al., “Aldehyde dehy-drogenase activity identi�es a population of human skeletalmuscle cells with highmyogenic capacities,”Molecularerapy,vol. 17, no. 11, pp. 1948–1958, 2009.

[26] E. Jean, D. Laoudj-Chenivesse, C. Notarnicola et al., “Aldehydedehydrogenase activity promotes survival of human muscleprecursor cells,” Journal of Cellular andMolecularMedicine, vol.15, no. 1, pp. 119–133, 2011.

[27] E. H. Huang, M. J. Hynes, T. Zhang et al., “Aldehyde dehydro-genase 1 is a marker for normal and malignant human colonicstem cells (SC) and tracks SC overpopulation during colontumorigenesis,” Cancer Research, vol. 69, no. 8, pp. 3382–3389,2009.

[28] C. Ginestier, M. H. Hur, E. Charafe-Jauffret et al., “ALDH1 is amarker of normal and malignant human mammary stem cellsand a predictor of poor clinical outcome,” Cell Stem Cell, vol. 1,no. 5, pp. 555–567, 2007.

[29] P. Marcato, C. A. Dean, D. Pan et al., “Aldehyde dehydrogenaseactivity of breast cancer stem cells is primarily due to isoformALDH1A3 and its expression is predictive of metastasis,” StemCells, vol. 29, no. 1, pp. 32–45, 2011.

[30] C. S. Sondergaard, D. A. Hess, D. J. Maxwell et al., “Humancord blood progenitors with high aldehyde dehydrogenaseactivity improve vascular density in amodel of acutemyocardialinfarction,” Journal of Translational Medicine, vol. 8, article 24,2010.

[31] B. J. Capoccia, D. L. Robson, K. D. Levac et al., “Revascular-ization of ischemic limbs aer transplantation of human bonemarrow cells with high aldehyde dehydrogenase activity,”Blood,vol. 113, no. 21, pp. 5340–5351, 2009.

[32] G. I. Bell, H. C. Broughton, K. D. Levac, D. A. Allan, A. Xeno-costas, and D. A. Hess, “Transplanted human bone marrowprogenitor subtypes stimulate endogenous islet regenerationand revascularization,” Stem Cells and Development, vol. 21, no.1, pp. 97–109, 2012.

[33] E. C. Perin, G. Silva, A. Gahremanpour et al., “A randomized,controlled study of autologous therapy with bone marrow-derived aldehyde dehydrogenase bright cells in patients withcritical limb ischemia,” Catheterization and CardiovascularInterventions, vol. 78, no. 7, pp. 1060–1067, 2011.

[34] E. C. Perin, G. V. Silva, Y. Zheng et al., “First randomized,controlled double-blind trial of transendocardial injection ofautologous aldehyde dehydrogenase-bright stem cells,” Amer-ican Heart Journal, vol. 161, no. 6, pp. 1078–11087, 2011.

[35] E. C. Perin, G. V. Silva, Y. Zheng et al., “Randomized, double-blind pilot study of transendocardial injection of autologousaldehyde dehydrogenase-bright stem cells in patients withischemic heart failure,” American Heart Journal, vol. 163, no. 3,pp. 415–421, 2012.

[36] K. Urbanek, D. Cesselli, M. Rota et al., “Stem cell niches inthe adult mouse heart,” Proceedings of the National Academyof Sciences of the United States of America, vol. 103, no. 24, pp.9226–9231, 2006.

[37] A. Meinhardt, A. Spicher, M. E. Roehrich, I. Glauche, P. Vogt,and G. Vassalli, “Immunohistochemical and �ow cytometricanalysis of long-term label-retaining cells in the adult heart,”Stem Cells and Development, vol. 20, no. 2, pp. 211–222, 2011.

[38] C. Castaldo, F. Di Meglio, D. Nurzynska et al., “CD117-positivecells in adult human heart are localized in the subepicardium,and their activation is associatedwith laminin-1 and𝛼𝛼6 integrinexpression,” Stem Cells, vol. 26, no. 7, pp. 1723–1731, 2008.

[39] M. Baddoo, K. Hill, R. Wilkinson et al., “Characterization ofmesenchymal stem cells isolated from murine bone marrow bynegative selection,” Journal of Cellular Biochemistry, vol. 89, no.6, pp. 1235–1249, 2003.

[40] H. Qian, K. Le Blanc, andM. Sigvardsson, “Primary mesenchy-mal stem and progenitor cells from bone marrow lack expres-sion of CD44 protein,”e Journal of Biological Chemistry, vol.287, no. 31, pp. 25795–25807, 2012.


[41] M. Al-Hajj, M. S. Wicha, A. Benito-Hernandez, S. J. Morrison,and M. F. Clarke, “Prospective identi�cation of tumorigenicbreast cancer cells,” Proceedings of the National Academy ofSciences of the United States of America, vol. 100, no. 7, pp.3983–3988, 2003.

[42] C. Chen, Y. Wei, M. Hummel et al., “Evidence for epithelial-mesenchymal transition in cancer stem cells of head and necksquamous cell carcinoma,” PLoS ONE, vol. 6, no. 1, Article IDe16466, 2011.

[43] Y. Alnouti and C. D. Klaassen, “Tissue distribution, ontogeny,and regulation of aldehyde dehydrogenase (Aldh) enzymesmRNA by prototypical microsomal enzyme inducers in mice,”Toxicological Sciences, vol. 101, no. 1, pp. 51–64, 2008.

[44] H. Storrie White, L. Smith, T. Gentry, and A. E. Balber,“Mechanisms of action of human aldehyde dehydrogenasebright cells in therapy of cardiovascular diseases: expres-sion analysis of angiogenic factors and aldehyde dehydro-genaseisozymes,” Journal of Stem Cell Research. In press,http://dx.doi.org/10.4172/2157-7633.S1-001.

[45] B. P. Levi, Ö. H. Yilmaz, G. Duester, and S. J. Morrison,“Aldehyde dehydrogenase 1a1 is dispensable for stem cellfunction in the mouse hematopoietic and nervous systems,”Blood, vol. 113, no. 8, pp. 1670–1680, 2009.

[46] P. Marcato, C. A. Dean, C. A. Giacomantonio, and P. W. K. Lee,“Aldehyde dehydrogenase its role as a cancer stem cell markercomes down to the speci�c isoform,” Cell Cycle, vol. 10, no. 9,pp. 1378–1384, 2011.

[47] J. S. Moreb, H. V. Baker, L. J. Chang et al., “ALDH isozymesdownregulation affects cell growth, cell motility and geneexpression in lung cancer cells,”Molecular Cancer, vol. 7, article87, 2008.

[48] G. R. Budas, M. H. Disatnik, and D. Mochly-Rosen, “Aldehydedehydrogenase 2 in cardiac protection: a new therapeutictarget?” Trends in Cardiovascular Medicine, vol. 19, no. 5, pp.158–164, 2009.

[49] J. I. Garaycoechea, G. P. Crossan, F. Langevin, M. Daly, M. J.Arends, and K. J. Patel, “Genotoxic consequences of endoge-nous aldehydes on mouse haematopoietic stem cell function,”Nature, vol. 489, no. 7417, pp. 571–575, 2012.

[50] M. Nagano, K. Kimura, T. Yamashita et al., “Hypoxia responsivemesenchymal stem cells derived from human umbilical cordblood are effective for bone repair,” StemCells andDevelopment,vol. 19, no. 8, pp. 1195–1210, 2010.

[51] E. Buchdunger, C. L. Cioffi, N. Law et al., “Abl protein-tyrosinekinase inhibitor STI571 inhibits in vitro signal transductionmediated by c-Kit and platelet-derived growth factor receptors,”Journal of Pharmacology and Experimental erapeutics, vol.295, no. 1, pp. 139–145, 2000.

[52] M. Crisan, S. Yap, L. Casteilla et al., “A perivascular origin formesenchymal stem cells in multiple human organs,” Cell StemCell, vol. 3, no. 3, pp. 301–313, 2008.

[53] P. Campagnolo, D. Cesselli, A. Al Haj Zen et al., “Human adultvena saphena contains perivascular progenitor cells endowedwith clonogenic and proangiogenic potential,” Circulation, vol.121, no. 15, pp. 1735–1745, 2010.

[54] R. Katare, F. Riu, K. Mitchell et al., “Transplantation of humanpericyte progenitor cells improves the repair of infarcted heartthrough activation of an angiogenic program involving micro-RNA-132,” Circulation Research, vol. 109, no. 8, pp. 894–906,2011.

[55] B. T. Estes, A. W. Wu, R. W. Storms, and F. Guilak, “Extendedpassaging, but not aldehyde dehydrogenase activity, increases

the chondrogenic potential of human adipose-derived adultstem cells,” Journal of Cellular Physiology, vol. 209, no. 3, pp.987–995, 2006.

[56] M. Petrini, S. Pacini, L. Trombi et al., “Identi�cation andpuri�cation of mesodermal progenitor cells from human adultbone marrow,” Stem Cells and Development, vol. 18, no. 6, pp.857–866, 2009.

Submit your manuscripts athttp://www.hindawi.com

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation http://www.hindawi.com

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttp://www.hindawi.com

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research


International Journal of

Microbiology

3ftfbsdi sujdmf $ibsbdufsj[bujpopg$bsejbd …downloads.hindawi.com/journals/bmri/2013/503047.pdf ·...

Documents